Ashwin Machanavajjhala

Ashwin Machanavajjhala is a Professor in the Department of Computer Science at Duke University's Pratt School of Engineering. With over 166 publications spanning from 2001 to 2025, his research has significantly impacted the fields of differential privacy, database systems, and data security. His recent work focuses on practical applications of differential privacy for government data releases, particularly for the US Census Bureau.

Machanavajjhala's research primarily centers on differential privacy, with significant contributions to database systems, privacy-preserving data analysis, and statistical disclosure control. His work bridges theoretical foundations with real-world applications, particularly in government statistics and census data protection. He has developed numerous frameworks and algorithms including DPXPlain for explaining differentially private query results, PreFair for generating fair synthetic data, and various components of the US Census Bureau's disclosure avoidance system. His research demonstrates a consistent trajectory from theoretical privacy mechanisms toward practical implementations that balance privacy guarantees with data utility.

His recent publications reveal a strong focus on applying differential privacy to census data (SafeTab, PHSafe), developing methods for explaining private query results (DPXPlain), addressing fairness in private data analysis (PreFair), and exploring privacy applications in blockchain technology. His work shows increasing engagement with government agencies, particularly the US Census Bureau, where his research has directly informed disclosure avoidance systems for the 2020 Census.

Machanavajjhala has advised numerous PhD students who have become prominent researchers in privacy and databases, including Ryan McKenna, Xi He, Yuchao Tao, and David Pujol. His collaborative network includes leading researchers from major institutions, with frequent collaborations with Gerome Miklau, Michael Hay, and Daniel Kifer. His research has been consistently funded by major grants supporting privacy-preserving data analysis.

He leads research on the Tumult Analytics framework, a robust and scalable differential privacy system, and has been instrumental in developing privacy technologies for the US Census Bureau's 2020 data release. His work demonstrates a commitment to making differential privacy practical for real-world statistical agencies and data providers.